Introduction:
Approximately 7.5% of adult Australians have diabetes mellitus (DM); 25% of hospital inpatients have DM. Inpatient hypoglycaemia occurs frequently but its significance is under recognised.
Aims:
To determine the incidence of and predictors for inpatient hypoglycemia, and the association of hypoglycaemia with mortality and hospital length of stay (LOS) in an Australian tertiary hospital.
Methods:
All patients admitted to Nepean Hospital over 12 months from July 2018 to June 2019 who experienced an inpatient hypoglycaemic event (HE), defined by capillary blood glucose level<4mmol/L, were included. HE patients were assigned to two groups: those with a diagnosis of DM and those without. The control group comprised randomly selected inpatients admitted over the same time period with a DM diagnosis but no HE.
Data extracted from electronic medical records (eMR) included: age, gender, length of stay (LOS), diabetes type, kidney injury, serum albumin, nutritional state dietitian assessed), presence of active malignancy, HbA1c, BMI, hypoglycaemic medications, hypoglycaemic awareness, medical team documentation of HE, in-hospital mortality and 12-month mortality.
Results:
251 hypoglycaemic events were identified: 202 (80%) in patients with DM and 49 (20%) without diabetes. 200 patients with DM without a HE formed the control group.
Risk factors for a HE included being malnourished (p=0.03), basal-bolus insulin use (p<0.01), mixed insulin use (p=0.01) and sulphonylurea use (p=0.03).
In patients with DM, comparing those with and without HE: median LOS was three-fold higher (12 versus 4 days), mean in-patient mortality was 3 fold higher (10% versus 3%) and mean 12-month mortality was 4-fold higher (22.5% versus 5.5%). Patients without DM with HE had a 34% 12-month mortality.
Conclusion:
Inpatient hypoglycaemia is common and is associated with a trebling of inpatient mortality, hospital length of stay and a 4-fold increase in 12-month mortality. Further research into preventative measures is urgently needed.